To be able to build scripts for modelling it is necessary to have access to the Z selection buffer, so you know which verts/edges/faces are not covered by others. Implementing such a Z-buffer through Python is probably way too slow (although I never tried). Whats more, why do this through Python when the framework already exists.
EdgeTune currently implements culling through checking normals. This is a very limited approach. Other plugins in the tink-tank here at EWOCprojects will never see the light of day without Z-buffer access.
Any supporters ?
Don’t feed your ego, feed the duckies